development

Plugin development from marketplace gb-plugins-marketplace in gautam-achieveai/ClaudePlugins. Commands read from the repo's manifest — 4 days ago.

Skills (15)

• agile-development

  Use when a large feature spans multiple independent workstreams or the user asks for Agile team mode, autonomous end-to-end delivery, or sustained single-prompt development.

• autonomous-design

  Internal helper. Load only when explicitly named by another skill or agent.

• brainstorming

  Internal helper. Load only when explicitly named by another skill or agent.

• compound-learning

  Use when a task has just finished and something non-obvious was learned, or at plan time to read back past lessons before designing or writing tasks.

• course-correction

  Use when a fix stacks on a fix, the same failure survives three attempts, the diff keeps spreading, or you cannot explain why a line is there — before trying anything else.

• draft-bug

  Internal helper. Load only when explicitly named by another skill or agent.

• draft-feature

  Internal helper. Load only when explicitly named by another skill or agent.

• draft-work-item

  Router that turns rough requirements into a well-structured work item and posts it to GitHub (issue) or Azure DevOps (work item). Detects the backend automatically from the git remote, classifies the intent, then delegates to a deep sub-skill — draft-feature for new features/user stories, draft-bug

• implement

  Internal helper. Load only when explicitly named by another skill or agent.

• receiving-code-review

  Internal helper. Load only when explicitly named by another skill or agent.

• scenario-driven-development

  Use when implementing a feature, fixing a bug, or changing behavior and deciding how to prove it works — the default testing mode for the development plugin.

• subagent-driven-development

  Internal helper. Load only when explicitly named by another skill or agent.

• test-driven-development

  Opt-in. Use only when the user explicitly asks for TDD or test-first development. The default testing mode is development:scenario-driven-development.

• verification-before-completion

  Internal helper. Load only when explicitly named by another skill or agent.

• work-on

  Provider-agnostic autonomous two-phase development workflow driven by a work item — a GitHub issue or an Azure DevOps work item. First run: analyzes the problem using all available tools (codebase, logs, database, observability, CI), designs a solution, creates a plan, posts it to the work item, and

Slash commands (2)

• /development:brainstorm

  Design before code - explore requirements, propose approaches, get approval before implementation

• /development:receive-review

  Triage and act on code review feedback — fix by severity/cost, push back on wrong or misaligned suggestions instead of blindly implementing

Subagents (5)

• architect

  Use this agent when a large feature or Agile workstream needs the smallest testable architecture, clear component boundaries, or a thin end-to-end slice before implementation. Typical triggers include decomposing a feature, defining interfaces, and simplifying a proposed design. See "When to invoke"

• blind-spot-detector

  Internal subagent. Invoke only when explicitly dispatched by an orchestrator skill.

• critic

  Use this agent when an Architect's substantial feature proposal or a Manual Tester's first runnable scenario needs adversarial review for blind spots, unnecessary complexity, missing consumers, or weak evidence. Typical triggers include challenging a thin-slice design, checking manual reachability,

• developer

  Use this agent when an Agile thin slice needs implementation, manual-test entry points, observed-defect regression tests, and behavioral tests for material acceptance contracts. Typical triggers include building the first tiny runnable path with the Manual Tester, debugging an observed failure, and

• manual-tester

  Use this agent when an Agile feature needs the earliest manually testable slice, continuous hands-on scenario testing, defect discovery, acceptance-contract mapping, or a final coverage-gap review. Typical triggers include pairing with the Developer on an entry point, testing a partial user flow, re

Hooks (3)

• PreToolUse

  node "${CLAUDE_PLUGIN_ROOT}/hooks/guard-bypass.mjs"

• PostToolUse

  node "${CLAUDE_PLUGIN_ROOT}/hooks/stop-gate.mjs" log

• Stop

  node "${CLAUDE_PLUGIN_ROOT}/hooks/stop-gate.mjs" stop

Similar plugins